Previous Research

Incremental Calcium Release and the IP3 Receptor

As an undergraduate, I studied the incremental release behavior of the IP3 receptor in smooth muscle cells under the guidance of Dr. Robert Phair. I developed a system for computer aided navigation through parameter space in an attempt to understand the ability of computational models to represent experimentally observed data. (more...)

A view into parameter space: the effect of changing one parameter of the kinetic model on the release of calcium by the IP3 receptor.

Template Based Segmentation of Medical Images: XBS

I developed a software tool for rapid segmentation of multiple image datasets using surface templates. At its core is a visualization engine which allows medical images to be viewed in a 3D coordinate system. The user interactively deforms the embedded surfaces so that they correctly segment a region of interest in the images. Multiple non-coplanar image sets can be segmented concurrently by computing the intersection of the slice locations with the anatomical surfaces. (SPIE 2000)		B-spline surface models of the heart chamber's endocardial wall embedded in MR images.

Respiratory Motion of the Heart

Magnetic resonance imaging (MRI) may one day provide a complete picture of coronary artery disease during a single non-invasive examination. However, cardiac and respiratory motion during the image acquisition procedure can lead to artifacts which degrade image quality. Improvements in image resolution and quality, and a reduction in imaging time are required for MR coronary imaging to become clinically useful.		Post-mortem 3D MRI of a human heart showing excellent coronary delineation in the absence of motion.
One way to reduce imaging times is to use motion correction during an MR exam. However, before implementing these techniques, we must first have a thorough understanding of how the coronary arteries move - during a cardiac contraction, and during a normal breathing cycle. I spent most of 2001 at INRIA in Sophia Antipolis, working with the Chir medical robotics group. Together, we developed methods for reconstructing the 3D structure and motion of coronary arteries from biplane x-ray angiograms. (SPIE 2002, TMI 2003)		A sample X-ray coronary angiogram.
I returned to the NIH to finish my PhD in the lab of Dr. Elliot McVeigh. Together with Dr. Jon Resar, we acquired angiogram movies from patients at Johns Hopkins. I reconstructed the motion of the heart, and developed a method for separating the cardiac and respiratory motion fields from one another (TMI 2004, SPIE 2004). After much analysis, the result was a PhD thesis entitled: "Respiratory motion of the heart: Implications for Magnetic Resonance Coronary Angiography." ( SCMR 2004, ISMRM 2004, MEDPHYS 2005, ISMRM 2005, TMI 2006)		3D model of the coronary arteries reconstructed from two x-ray projections.